Modelpublic

BojanaBas/Mistral-7B-Instruct-v0.2-pqa-10

sourceHugging Faceagpl-3.0updated 2y agoView on Hugging Face

0likes31downloads

Model Card

Training procedure

The following bitsandbytes quantization config was used during training:

—quant_method: bitsandbytes
—loadin_8bit: False
—loadin_4bit: True
—llmint8threshold: 6.0
—llmint8skip_modules: None
—llmint8enablefp32cpu_offload: False
—llmint8hasfp16weight: False
—bnb4bitquant_type: fp4
—bnb4bitusedoublequant: False
—bnb4bitcompute_dtype: bfloat16
—bnb4bitquant_storage: uint8
—loadin4bit: True
—loadin8bit: False

Framework versions

—PEFT 0.5.0

arXiv paper

—https://arxiv.org/abs/2407.05015